Hello, I am chasing knock count going crazy after car runs for a short time, resulting in retarded timing and lack of power. I believe I have narrowed it down to the ECM. I have had my ECM rebuilt twice now which has not fixed the issue. I have also tried several PROMs including a brand new PROM with same results. Any ideas are much appreciated. I believe I am looking for a good 92' Lt1 ECM 16159278. Thank you.

 

Did you or a previous owner install new knock sensors? If so, you might want to check their torque setting. I believe they are to be torqued to about 14 lb/ft or so. If you go any tighter than that, they reportedly start "false knocking". This is also the case with other name brands (Honda, Nissan, etc.)

Get under the car and see if the are new. May not be a bad idea to replace them either way.

Ric

 

Check the sensors and check the wiring and the connectors/terminals as well. Lot of heat in that area and also age will take a toll.

 

And only use GM knock sensors. The aftermarket stuff is often junk.

My recent experience with false knock. Different ECM but similar diagnostics.

https://www.thirdgen.org/forums/diy-...tard-real.html

 

Just for background information, are there any modifications to the engine mechanicals?

 

Yep... roller rockers, headers, etc... they'll generate noises and fool the KS. Which is what was happening to me.

 

I'm today's global economic landscape, a GM sensor may not be related to GM at all. I bought a Wells Electronics TPS from Rock Auto and noticed that the GM counterpart had the same picture and description. The Wells unit was 30% cheaper, so I got it. Over a year now... No issues.

Unfortunately, it is not the 90s anymore.

Damn Facebook, tiktok, and every other social media junkyard. 🤬

 

Just engineering curiosity here. Your last sentence is past tense. What did you do to deal with the altered noise signature of the engine?

 

It's in that link I posted earlier to the thirdgen.org site...

 

Please see attached video. This is me sitting in the garage with my cheapy scan tool. Spark advance -14 at idle and even higher if I rev the engine? Is this normal?

 

I'm assuming the minus sign indicates BTDC. What year car?

Looking at few different bins up to 1996, it seems like you should be at least between 20-25 BTDC at idle and ~40 kPa.

Does you scan tool see knock retard too? Are you getting any?

 

It's a 96 LT-4. Apparently, some scanner read with a "-", but doesn't necessarily mean the ECM is taking running. These scanner have a+ when the running is retarded.

 

Ok, yeah I think you should be getting mid 20's BTDC.

 

How in the world is my car at 14 degrees then at idle? What's Max at WOT? Would this have anything to do with the higher compression on the LT-4?

 

All stock internals and exhaust.

 

Okay, thanks.

Additional (engineering) questions: What is driving your thought that the ECM is suspect? Also, have the knock sensors been removed in the past, or have the connectors been unplugged multiple times?

 

I don't the software to read OBD-II (I can only see up to OBD-1 1995).

But I don't suspect they'd be much different since the engines are about the same.

Here's a 1995 Vette spark timing map, along with the idle speed spark correction tables. You'll get some amount of swing on the spark timing, but you should be about centered on the value in the main table.

 

When this problem first started, I changed both knock sensors with genuine GM parts torqued to 14 lb ft. All wiring from ECM to ks sensors check good with continuity,. I unplugged the ks in the block and connected my old ks sensors, just hanging (not installed) with a ground wire to each and still have the same results. I also used a 3500 ohm resistor in place of the ksensors and have the same results. I have also tried several PROM chips with no change. If I unplug both ks I get code 43. Not sure what else to do...

 

Edit... my question about ohming out the wires already got answered.

 

Okay, some comments and additional questions. Years ago a coworker and I worked on a car that GM had bought back from the customer due to somewhat similar issues as you mention (the dealership spent a lot of time debugging the issue, and replaced the sensors, memcal, and even the ECM, but the false knock issue kept reoccurring). Because the car was now GM property and my coworker and I were making substantially less than the $100/hr dealer shop rate, we could take our time and check a bunch more details without worrying about the customer wanting his car back soon, or complaining about altered wiring under the hood. On THIS one vehicle we found the problem to be the electrical integrity of the knock sensor connector. A Delco Remy engineer we consulted mentioned that these connectors are only "rated" for IIRC four plugs/unplugs due to the material in the connector interface (the material was there to improve the conductivity, but was fragile if manipulated too many times). We replaced the sensor harness connector, and the fault went away. Is this the same issue with your car? I don't know, this is just an FYI to be cautious with the handling of the connector.

Now the questions: With the resistors substituted instead of the sensors, are you seeing a code or false knock? The ESC circuitry in the ECM has frequency and time filters built into it, so I would expect it to not react to having just load resistors hooked to the input. (But, occasionally a fault issue can prove me wrong. )

Second question: Do you have access to an oscilloscope to monitor the sensor signal lines?